Associations between a preference for violent electronic games and adolescents’ self-perceptions of problem behaviors and emotions were examined. It was predicted that a preference for violent games would be associated with negative externalizing characteristics, in particular aggressive emotions and behaviors, on the Youth Self-Report (YSR), a standardized self-report measure of adolescent problem behaviors. Thirty-two 11- through 15-year-olds (17 girls) listed and categorized favorite electronic games into one of six predetermined categories and completed the YSR. MANOVA revealed significant relationships between a preference for violent games and the Thought Problems subscale (P < .01) and YSR Total Score (P < .05), with trends noted for the Internalizing (P < .06) and Anxious-Depressed (P < .08) subscales. Expected relationships with externalizing behaviors, including aggression, were not found. However, across all YSR subscales, children with higher preference for violent games had more clinically significant elevations than those with low preference for violent games. On the Total Problems subscale, of the eight children receiving scores in the clinically significant range, six were in the High preference group (three boys and three girls). The failure to find the expected relationships between a preference for violent games and aggressive, externalizing behaviors is puzzling. It is possible that individuals with a preference for violent games may have high exposure to all forms of media violence. Their perceptions of their own behavior, in comparison, may not seem sufficiently aggressive to justify endorsement of problems in this area. Or, playing violent electronic games may promote a disconnection between the emotions normally associated with violence and violent acts. These explanations are consistent with a desensitization model where exposure to media violence decreases sensitivity to aggression. Aggr. Behav. 28:134–144, 2002. © 2002 Wiley-Liss, Inc.